End-of-Life Care Preferences and Needs: Perceptions of Patients with Chronic Kidney Disease

Discussion

The illness trajectory of patients dying with CKD differs from that of patients with cancer (12–14). There are also unique end-of-life issues, such as the withdrawal of dialysis, that are specific to these patients. The existence of different pathways to death has important implications for health care delivery; end-of-life issues that are important to CKD patients may differ from patients dying with other illnesses. Identifying these elements and preferences for care is essential for optimizing CKD care near the end of life.

Communication of prognosis and discussions related to planning for future death are lacking in the routine care of CKD patients. The vast majority of these patients did not appreciate the likelihood of clinical deterioration over the next year. Given the mean age of 68 years, these patients have a 50% survival rate of less than 3 years (1). These patients, however, feel it is important that they receive prognostic information and plan for future death. Most patients reported that they think about their future health and that they are comfortable talking about end-of-life care issues with both family and renal staff. Even those less comfortable with having these discussions still feel they needed to occur. This is consistent with a Canadian study where 97% of 100 general nephrology patients during their first visit to a nephrologist wanted detailed information about survival before obtaining informed consent for chronic dialysis and felt that the physician should provide this information without being prompted (15). Sentinel events (hospitalizations, acute illnesses) present opportunities to engage in end-of-life care discussions (16). Answering yes to a simple question “Would you be surprised if this patient died with the next year?” should also prompt the nephrology team to initiate these discussions (17).

Although patients want to primarily involve family and friends in end-of-life discussions, nephrologists and family physicians are considered integral to the process, and a substantial proportion of patients reported that they would rely on their nephrologist to make medical decisions on their behalf should they become incompetent to make decisions for themselves. Unfortunately, neither physicians nor family members are accurate in predicting patients' desires about life-sustaining treatments, including wishes for ongoing dialysis. In a study of 398 dialysis patients, only 44% of physicians correctly predicted the patient's current preference for cardiopulmonary resuscitation (CPR), 47% their wish for dialysis in a severely demented state, and 43% their wish for dialysis if they had terminal cancer. The corresponding figures for family members were 50%, 44%, and 47% (18). The majority of nephrologists do not feel well prepared to make end-of-life decisions for their patients (19). Physicians are often hesitant about having these conversations with their patients because of concerns about destroying patients' hope. Evidence is emerging, however, that for many dialysis patients, the opposite is true and that focusing on immediate clinical care without attention to future goals is a barrier to maintaining hope (20). Unfortunately, over 90% of patients in this study reported having had no discussion about prognosis with a doctor. Less than 10% have engaged in any form of end-of-life care discussion with their nephrologist, and even fewer with other healthcare professionals. In fact, most of these patients have no advance directive. Unless advance care planning that outlines goals for end-of-life care is systematically integrated into the care of ESRD patients, their end-of-life care needs are unlikely to be adequately met.

Patients clearly need enhanced education with respect to end-of-life care issues. Most patients in this study prefer care focused on decreasing pain and suffering; only 18% of patients preferred clinical care that focuses on extending life, even if at the expense of prolonging pain and discomfort. Unfortunately, dialysis patients typically have poor knowledge of their chances of survival after CPR (21,22). Despite the fact that CPR rarely extends survival for dialysis patients (23–26), over a third of these patients would want to be resuscitated.

Many of these patients regret their decision of dialysis over conservative care. Although symptoms or a history of depression were not elicited, given the high prevalence of depression in CKD, depression could have been an important driver of some patient responses. Regardless, these patients reported that their decision to be treated with dialysis reflects physicians' and family members' preferences rather than personal choice. This highlights the need to reevaluate decision-making around the initiation of dialysis and involving patients in discussions about prognosis and goals of care (27). New tools for facilitating advance care planning in CKD are available, and an approach that emphasizes a relational, patient-centered process that focuses on broader goals of care has been developed (28,29). The lack of knowledge of palliative care will obviously hinder patients' appreciation for the supportive care services that could be made available to them, especially if they choose conservative care.

Not surprisingly, symptom management was clearly a priority for patients who felt that this was a responsibility of their renal care team. Unfortunately, pain in CKD is both under-recognized (30) and undertreated (31). The nephrologist and nephrology nurse are also perceived as important for providing emotional, social, and spiritual support. This supportive role was once recognized as an integral component of medicine. However, with the advent of modern technology, physicians have departed from this role, with emotional and spiritual support falling largely to nursing staff and allied health professionals, such as social workers, spiritual counselors, and pastoral care. The vast majority of these patients do not rely on support counselors or spiritual advisors for their social and emotional support. This may reflect a lack of presence of these members of the multidisciplinary team in our renal program. How open patients would be to these allied health professionals if they were more readily available is not clear.

Previous studies involving palliative patients suggest a preference for dying at home followed by inpatient hospice care (32). However, more recently, a Canadian study of hospitalized patients with cancer and end-stage medical conditions suggested that only half of all patients and family caregivers report a preference for a home death (33). Our data confirm that although approximately one fourth of patients preferred to die in hospital, 65% of patients preferred to die either at home or in a hospice. Given the poor self-reported knowledge of palliative care and hospice, this proportion of patients may increase as they are made aware of the potential benefits associated with palliative and hospice care. Unfortunately, the vast majority of dialysis patients die in acute care facilities, without accessing palliative care services (7). End-of-life care policies and resources should be directed toward ensuring that patients die in their location of choice, which includes not only home but within-institutional settings, such as in-patient hospice.

There are several limitations of this study, some of which are intrinsic to the use of questionnaires. Responses can be difficult to interpret because patients' understanding of questions and reasons why they answer as they do are not sought directly. These findings are limited predominantly to white, English-speaking patients. Our data, however, highlight end-of-life clinical practices that do not meet the needs for patients with CKD. Changing attitudes and improving the knowledge of renal health care professionals and patients about palliative care are going to be the key to improving renal supportive care. We also need to define the role that various members of the multidisciplinary team will play in the provision of end-of-life care, such as symptom management, advance care planning, and psychosocial and spiritual support. This study, however, is a first step in identifying gaps between current end-of-life care practice and patients' preferences and will hopefully help prioritize and guide future innovation in renal end-of-life care policy. Recommendations for end-of-life care practices that should be integrated into CKD care are outlined in Table 6.